Traumatic injury of the spinal cord results in permanent functional impairment. Most of the deficits associated with spinal cord injury result from the loss of axons that are damaged in the central nervous system (CNS). Similarly, other diseases of the CNS are associated with axonal loss and retraction, such as stroke, HIV dementia, prion diseases, Parkinson's disease, Alzheimer's disease, multiple sclerosis and glaucoma. Common to all of these diseases is the loss of axonal connections with their targets, and the ability to stimulate growth of axons from the affected or diseased neuronal population would improve recovery of lost neurological functions. For example, following a white matter stroke, axons are damaged and lost, even though the neuronal cell bodies are alive. Treatments that are effective in eliciting sprouting from injured axons are equally effective in treating some types of stroke (Boston life sciences, Sep. 6, 2000 Press release). Similarly, although the following discussion will generally relate to delivery of Rho antagonists, etc. to a traumatically damaged nervous system, this invention also pertains to damage from unknown causes, such as during multiple sclerosis, HIV dementia, Parkinson's disease, Alzheimer's disease, prion diseases or other diseases of the CNS were axons are damaged in the CNS environment.
It has been proposed to use various agents to stimulate regeneration of cut axons, i.e. nerve lesions. Please see for example Canadian patent application Nos. 2,304,981 (McKerracher et al) and 2,300,878 (Strittmatter). These document propose the use of known Rho antagonists such as for example C3, chimeric C3 proteins, etc. (see blow) as well as substances selected from among known trans-4-amino(alkyl)-1-pyridylcarbamoylcyclohexane compounds (also see below) or Rho kinase inhibitors for use in the regeneration of axons.
Several major advances in our understanding of axon regeneration have led to the ability to stimulate some axon regeneration and functional repair in animal models of spinal cord injury. In the 1980's experiments by Aguayo and colleagues to use peripheral nerve grafts that were inserted into the brain or spinal cord showed that CNS neurons have the capacity to regrow, and these studies highlighted that diverse classes of CNS neurons have the potential to regenerate when given a permissive growth environment (Aguayo, et al. (1981) J Exp Biol. 95:231-40). However, this technique cannot be used to rewire the complex circuitry of the CNS. Another major advance in our understanding of axon regeneration in the central nervous system was the discovery by Schwab and colleagues that the CNS environment did not simply lack growth promoting molecules, but that growth inhibitory molecules existed to block axon growth (Schwab, et al. (1993) Annu. Rev. Neurosci. 16:565-595). Long distance regeneration in the CNS by blocking growth inhibitory molecules with antibodies was first achieved in juvenile rats by neutralization of inhibitory protein activity with the IN-1 antibody in spinal cord (Schnell and Schwab (1990) Nature. 343:269-272) and optic nerve (Weibel, et al. (1994) Brain Res. 642:259-266). However, this technique suffers from the problem that only a single growth inhibitory protein is targeted, and delivery by the application of hybridoma cells or by infusing antibodies with pumps. There have been investigations on the use of growth factors to promote regeneration in the CNS, some with notable success (Ramer, et al. (2000) Nature. 403:312-316, Liu, et al. (1999) J Neurosci. 19:4370-87, Blesch, et al. (1999) J Neurosci. 19:3556-66). Typically infusion pumps or gene therapy techniques are used to deliver growth factors to injured neurons. In general, trophic factors do not stimulate long distance regeneration, but stimulate more of a local sprouting response (Schnell, et al. (1994) Nature. 367:170-173, Mansour-Robaey, et al. (1994) Proc. Natl. Acad. Sci. 91:1632-1636).
A more recent advance is the demonstration that increasing the intrinsic growth capacity of neurons is sufficient to allow axon regeneration in the CNS, and that neurons primed for regeneration with neurotrophins, a conditioning lesion, or treatment with Rho antagonists have a better chance to grow on inhibitory substrates (Neumann (1999) Neuron. 23:83-91, Cai, et al. (1999) Neuron. 22:89-101, Lehmann, et al. (1999) J. Neurosci. 19:7537-7547). Targeting intracellular signalling mechanisms is likely to be the most efficient way to promote axon regeneration, and it has been found that Rho antagonists are able to stimulate regeneration in the optic nerve of adult rats (Lehmann et al (1999) IBID). However, preliminary experiments to apply Rho antagonists to the injured spinal cord were not successful. It is believed that the infused protein was not sufficiently retained at the injury site, either by syringe application or the use of Gelfoam. This suggested that the delivery of compounds that act with low affinity (compared to high affinity neurotrophins) posed unique problems in delivery. As shall be discussed in greater detail below the present invention relates to a tissue-adhesive delivery system whereby the Rho antagonist is added to the adhesive solution before application of the solution with a syringe, and polymerization of the adhesive within the lesion cavity in the CNS.
While neurons in the peripheral nervous system regenerate naturally, there are many techniques used to enhance and help the repair process. Most of these techniques are not aimed at stimulating the rate of axonal regeneration, but in helping to guide axons back towards their target regions. For example, severed nerve are sewn or glued together with a fibrin glue enhance the repair process. While the following discussion will generally relate or be directed at repair in the CNS, the techniques described herein may be extended to use in PNS repair. Treatment with Rho antagonists in the adhesive delivery system could be used to enhance the rate of axon growth in the PNS. This is first use of Rho antagonists in the PNS.
Growth inhibitory proteins cause growth cone collapse (Li, et al. (1996) J. Neurosci. Res. 46:404-414, Fan, et al. (1993) J. Cell Biol. 121:867-878) and it has become clear that GTPases of the Rho family that comprise Rho, Rac and Cdc42 are intracellular regulators of growth cone collapse (Lehmann, et al. (1999) J. Neurosci. 19:7537-7547, Tigyi, et al. (1996) Journal of Neurochemistry. 66:537-548, Kuhn, et al. (1999) J. Neurosci. 19:1965-1975, Jin and Strittmatter (1997) J. Neurosci. 17:6256-6263). These small GTPases exist in inactive (GDP) and active (GTP) forms, and the cycling between active GTP-bound and inactive GDP-bound states is tightly regulated. The guanine nucleotide exchange factors (GEFs) accelerate the release of GDP, thereby facilitating GTP binding. The GTPase activating proteins (GAPs) catalyze GTP hydrolysis and conversion of the inactive form. The GDP dissociation inhibitors (GDIs) act to maintain Rho in a GDP-bound form. GEFs for Rho all have a domain homologous with the Db1 oncoprotein, and more than 20 such proteins have been identified, including Tiam-1 which is highly expressed in brain (Zheng and Li (1999) J. Biol. Chem. 272:4671-4679, van Leeuwen, et al. (1997) J. Cell Biol. 139:797-807). Once in the active form, Rho GTPases typically stimulate ser/thr kinases, such as ROK (Rho kinase), PAK (p21-activated kinase) and downstream effectors that act on the cytoskeleton.
The Rho family members that regulate the cytoskeleton and motility include Rho, Rac and Cdc42 (Nobes and Hall (1995) Cell 1995 81:53-62). Rho is an important link between signaling through integrins and signaling cascades of trophic factors (Laudanna, et al. (1996) Science. 271:981-983, Hannigan, et al. (1996) Nature. 379:91-96, Kuhn, et al. (1998)J. Neurobiol. 37:524-540). Cdc42 is important for the regulation of filopodia (Nobes and Hall (1995) Cell 1995.81:53-62). Both Rac and Rho regulate growth cone motility and axon growth. In non-neuronal cells a hierarchy of signaling between Rho, Rac and Cdc42 exists (Hall (1996) Ann. Rev. Cell Biol. 10:31-54). In neurons Rac and Rho may have opposite effects (van Leeuwen, et al. (1997) J. Cell Biol. 139:797-807, Kozma, et al. (1997) Molec. Cell. Biol. 17:1201-1211). Activation of Rac stimulates outgrowth of neurites from NIE-115 neuroblastoma neurons whereas activation of Rho causes neurite retraction (van Leeuwen, et al. (1997) J. Cell Biol. 139:797-807, Albertinazzi, et al. (1998) J. Cell Biol. 142:815-825). In PC12 cells, dominant negative Rac disrupts neurite outgrowth in response to NGF (Hutchens, et al. (1997) Molec. Biol. Cell. 8:481-500, Daniels, et al. (1998) EMBO Journal. 17:754-764) whereas treatment of PC12 cells with lysophosphatidic acid (LPA), a mitogenic phospholipid that activates Rho, causes neurite retraction (Tigyi, et al. (1996) Journal of Neurochemistry. 66:537-548). The p21-activated kinase(PAK) is activated by Rac, and PAK can also induce PC12 cell neurite outgrowth (Daniels, et al. (1998) EMBO Journal. 17:754-764). It has been shown that inactivation of Rho is sufficient to promote PC12 cell neurite outgrowth on growth inhibitory substrates (Lehrnann, et al. (1999)J. Neurosci. 19:7537-7547). A recent study of activating and null mutations of Rho expressed in PC12 cells suggests that the differentiation state is an important parameter for the effect of Rho on neurite outgrowth, and that priming PC12 cells with NGF can alter the responsiveness to activating and null mutations (Sebok, et al. (1999) J. Neurochem. 73:949-960). This result is in agreement with the finding that priming neurons increases intracellular cAMP (Cai, et al. (1999) Neuron. 22:89-101), which can in turn influence the activation of Rho (Lang, et al. (1996) EMBO J. 15:510-519, Dong, et al. (1998) J. Biol. Chem. 273:22554-22562).
In primary neurons Rac and Rho regulate both dendrite and axon growth and cone morphology and collapse. By immunocytochemistry it has been demonstrated that Rho is concentrated in growth cones, and some colocalizes at sites of point contact (Renaudin, et al. (1998) J. Neurosci. Res. 55:458-471). Experiments with activating and dominant negative mutations have demonstrated that activation of Rac is important in maintaining a spread morphology after challenge with growth cone collapsing factors (Kuhn, et al. (1999) J. Neurosci. 19:1965-1975, Jin and Strittmatter (1997) J. Neurosci. 17:6256-6263). The activation of Rho induces growth cone collapse, and collapse can be prevented by treatment with Clostridium botulinum C3 exotransferase (hereinafter simply referred to as C3) (Tigyi, et al. (1996) Journal of Neurochemistry. 66:537-548, Jin and Strittmatter (1997) J. Neurosci. 17:6256-6263). C3 inactivates Rho by ADP-ribosylation and is fairly non-toxic to cells (Dillon and Feig (1995) Methods in Enzymology: Small GTPases and their regulators Part. B. 256:174-184).
An important downstream target of activated Rho is p160ROK, a Rho kinase (Kimura and Schubert (1992) Journal of Cell Biology. 116:777-783, Keino-Masu, et al. (1996) Cell. 87:175-185, Matsui, et al. (1996) EMBO J. 15:2208-2216, Matsui et al. (1998) J. Cell Biol. 140:647-657, Ishizaki (1997) FEBS Lett. 404:118-124). Among other effects, ROK phosphorylates myosin phosphatase to regulate actin-myosin based motility (Matsui, et al. (1996) EMBO J. 15:2208-2216) and regulates proteins of the ezrin family (Vaheri, et al. (1997) Curr. Opin. Cell Biol. 9:659-666), which are concentrated in neuronal growth cones (Goslin, et al. (1989) J. Cell Biol. 109:1621-1631). Activation of ROK also induces growth cone collapse, which can be prevented by the addition of the ROK inhibitor Y-27632 (Hirose, et al. (1998) J. Cell Biol. 141:1625-1636).
The above studies showed that Rho antagonists can stimulate regeneration in the CNS. It has been demonstrated that Rho kinase is an important downstream target of Rho signaling (Matsui, et al. (1996) EMBO J. 15:2208-2216, Bito (2000) Neuron. 26:431-441). Among other effects, inactivation of Rho kinase stimulates neurite outgrowth in tissue culture (Bito (2000) Neuron. 26:431-441) as does inactivation of Rho (Lehmann, et al. (1999) J. Neurosci. 19:7537-7547). Therefore, inactivation of Rho kinase should induce the same biological effects in vivo as inactivation of Rho.
The Rho kinase inhibitory Y-27632 compound is a trans-4-amino(alkyl)-1-pyridylcarbamoylcyclohexane compound; this compound is for example described in U.S. Pat. No. 4,997,834 the entire contents of which are incorporated herein by reference; this patent refers for example to compounds which may be selected from the group consisting of trans-4-aminomethyl-1-(4-pyridylcarbomoyl)cyclohexane, trans-4-aminomethyl-trans-1-methyl-1-(4-pyridylcarbamoyl)cyclohexane, trans-4-aminomethyl-cis-2-methyl-1-(4-pyridylcarbamoyl)cyclohexane, trans-4-aminomethyl-1-(2-pyridylcarbamoyl)cyclohexane, trans-4-aminomethyl-1-(3-pyridylcarbamoyl)cyclohexane, trans-4-aminomethyl-1[(3-hydroxy-2-pyridylcarbamoyl)]cyclohexane, trans-4-aminomethyl-1-(3-methyl-4pyridylcarbamoyl)cyclohexane, 4-(trans-4-aminomethylcyclohexylcarboxamido)-2,6-dimethyl-pyridine-N-oxide, trans-4-aminomethyl-1-(2-methyl-4-pyridylcarbamoyl)cyclohexane, trans-4-(2-aminoethyl)-1-(4-pyridylcarbamoyl)cyclohexane, trans-4-(1-amino-1-methylethyl)-1-(4-pyridylcarbamoyl)cyclohexane, trans-4-(1-aminopropyl)-1-(4-pyridylcarbamoyl)cyclohexane, and pharmaceutically acceptable acid addition salts thereof.
Please also see also Ishizali et al. 2000. Molecular Pharmacology 57:976-983 3 which refers to Y-27632 in the dihydrochloride form as well as to a related compound Y-30141, namely (R)-trans-4-(1-aminoethyl)-N-(1H-pyrrolo[2,3]pyridin-4-yl) cyclohexanecarboamide dihydrochloride. A patent application A Medicines comprising Rho kinase inhibitor has been submitted-(EPO 956 865 A1). This inhibitor has not been tested for efficacy in CNS injury, nor has the company who patented this compound discovered how it might to applied to a region of CNS injury in a kit form. Such a kit is provided in our invention. Please see also European Patent application no. 97934756.4; PCT/JP97/02793; International publication # WO 98/06433 (19.02.1998/07).
The compound Y-27632 has the following structure

The above structure is used herein in a pharmaceutically acceptable salt form (e.g. dihydrochloride salt).
The above mentioned related compound Y-30141 which may be exploited in accordance with the present invention has the following structure:

Again the above structure may also be used herein in a pharmaceutically acceptable salt form (e.g. dihydrochloride salt).
The compound (R)-(+)-trans-N-(4-pyridyl)-4-(1-aminoethyl)-cyclohexanecarboamide (Y-27632) inhibits Rho kinase at sub-micromolar concentrations (Uehata, et al. (1997) Nature. 389:990-994). Y-27632, made by a Yoshitoma, affects calcium sensitization of smooth muscles to affect hypertension. It was reported that the cellular target of Y-27632 is Rho-associated protein kinase, p160ROCK (Uehata, et al. (1997) Nature. 389:990-994, Somlyo (1997) Nature. 389:908-911).
Different methods have been used for local delivery of drugs in the CNS, however none of these methods have been developed as a kit with biological component that have proven effective in the promotion of the regeneration of injured axons. IN-1 is an antibody that promotes regeneration in the CNS. One method of delivery is the implantation of cells that secrete the active antibody (Schnell et al (1994) Nature 367:170). The use of fibrin adhesive for the delivery of IN-1 antibody was not found to be effective (Guest (1997) J. Neurosci. Res. 50:888-905). Another method is the use of pumps to infuse and deliver continuously over time compounds that stimulate regeneration. (Ramer, et al. 2000, Nature. 403:312-316, Verge, et al. 1995. Journal of Neuroscience. 15:2081-2096).
Fibrin adhesives per se have been used in studies of CNS regeneration. It has been used in replacement of sutures to graft peripheral nerves into the damaged CNS (Cheng, et al. (1996) Science. 273:510-513). A fibrin glue has also been used for the delivery of fibroplast growth factor (FGF) to damaged corticospinal neurons (Guest (1997) J. Neurosci. Res. 50:888-905). The use of fibrin glue plus FGF did not promote long distance regeneration.
Collagen per se has been tested for its ability to promote regeneration after injury (Joosten (1995) J. Neurosci. Res. 41:481-490). Collagen has also been used for the delivery of neurotrophins to injured corticospinal axons (Houweling (1998) Expt. Neurol. 153:49-59). Neither of the conditions was able to support long distance regeneration. In tissue culture, collagen gels can maintain gradients of small molecules important in axon guidance (Kennedy, et al. (1994) Cell. 78:425-435). Moreover, it had been reported that collagen gels by themselves could foster some axon regeneration after spinal cord injury (Joosten (1995) J. Neurosci. Res. 41:481-490).
Many different protein-based tissue adhesives exist Examples include collagen gels, fibrin tissue adhesives, matrigel, laminin networks, and adhesives based on a composition of basement membrane proteins that contain collagen: Perhaps the most popular are the fibrin adhesives.
Fibrin sealant has three basic components: fibrinogen concentrate, calcium chloride and thrombin. Other components can be added to affect the properties of the gel formation. Added components are used to modulate time it takes for the fibrin gel to form from the soluble components, the size of the protein network that is formed, the strength of the gel, and protease inhibitors slow down the removal of the gel after it is place in the body. Several different commercial preparations are available as kits. These include Tissucol/Tisseel, (Immuno AG, Vienna, now marketed by Baxter), Beriplast P, (Hoechst, West Germany), and Hemaseel (Hemacure Inc. Kirkland, Quebec).
To make a fibrin gel soluble thrombin and fibrinogen are mixed in the presence of calcium chloride. When the components mix, a fibrin adhesive gels is formed because the fibrinogen molecule is cleaved by thrombin to form fibrin monomers. The fibrin monomers spontaneously will polymerize to form a three-dimensional network of fibrin, a reaction that mimics the final common pathway of the clotting cascade, i.e. the conversion of fibrinogen to fibrin sealant. The key to the preparation of commercial preparations is to keep the fibrinogen and thrombin components separate until use, so that the polymerization can be controlled with the desired timing before or after application to the body.
Today such use of fibrin as a biological adhesive has been widely accepted and found application in many fields of surgery. HEMASEELJ or Tisseel VH are used as an adjunct to hemostasis in surgeries involving cardiopulmonary bypass and treatment of splenic injuries due to blunt or penetrating trauma to the abdomen, when control of bleeding by conventional surgical techniques, including suture, ligature and cautery is ineffective or impractical. The action of these fibrin gels is also used to stop bleeding in surgical procedures involving cardiopulmonary bypass and repair of the spleen. Tisseel VH has also been shown to be an effective sealant as an adjunct in the closure of colostomies.
Collagen gels have been used in tissue culture studies to main gradients of diffusible molecules. The use of collagen gels has permitted the identification and testing of neuronal guidance factors such as netrins (Kennedy, et al. (1994) Cell. 78:425-435). When collagen polymerized it forms a dense protein network. Therefore, like fibrin, it has the potential to act as a tissue adhesive. Moreover, collagen is easy to purify in large quantities.
There are many different types of collagens, and it is a major component of basement membranes in many different body tissues. The form of collagen often used for experimental studies in rodents is type IV collagen because it is easily purified from rat tails.
Not only is collagen a component of the basement membrane in the peripheral nervous system, but it is known that neurons express receptors for collagen. Receptors for collagens are receptors of the integrin class of proteins. One important collagen receptor expressed by neurons is the alpha1 beta1 receptor (McKerracher, et al. 1996. Molec. Neurobiol. 12:95-116); this receptor is involved in the promotion of neurite outgrowth. When PC12 cells, a neuronal cell line, are plated on collagen substrates in tissue culture, collagen helps promote neurite growth in an integrin-dependent fashion. The addition of anti-integrin antibodies block neurite outgrowth. Therefore, the ability of collagen, by itself, has been tested for its ability to promote axon regeneration after spinal cord injury. It was reported that collagen gels by themselves could foster some axon regeneration after spinal cord injury (Joosten (1995) J. Neurosci. Res. 41:481-490). However, the observed growth was more of a sprouting response with out any long distance regeneration past the glial scar and site of the lesion. In addition, collagen has been tested for its ability to promote regeneration after injury in conjunction with the delivery of neurotrophins to injured corticospinal axons (Houweling (1998) Expt. Neurol. 153:49-59). This treatment was not able to support long distance regeneration, although the treated animals had a better sprouting response than the controls.
It would be advantageous to have a means for the direct delivery to and maintenance at a lesion site of an agent able to facilitate axon growth at the lesion site.